Sie sind auf Seite 1von 7

American Journal of Epidemiology Copyright O 1997 by The Johns Hopkins University School of Hygiene and Public Health All

rights reserved

Vol 146, No. 11

Printed In USA.

Lactation History and Breast Cancer Risk

Jo L. Freudenheim,1 James R. Marshall,2 John E. Vena,1 Kirsten B. Moysich,1 Paola Murti,1 Rosemary Laughlin,1 Takuma Nemoto,3 and Saxon Graham1 Lifetime lactation in relation to breast cancer risk was examined in a case-control study in two counties in western New York. Cases were women age 40 years and over with incident, primary, histologicalty confirmed breast cancer. Controls were age- and county-frequency matched, selected from New York State driver's license records (for those less than age 65 years) and from Health Care Finance Administration Records (for those age 65 or more). Included were women with at least one livebirth (253 premenopausal and 367 postmenopausal cases and 266 premenopausal and 427 postmenopausal controls). Breast cancer risk was very weakly associated with long duration of lactation among premenopausal women; the odds ratio for at least 20 months lifetime lactation was 0.50 <95% confidence interval 0.21-1.12). Among postmenopausal women, the protective effect of lactation was restricted to women with first lactation before age 25 years (odds ratio = 0.67, 95% confidence interval 0.46-0.95). However, age at first birth was highly correlated with age at first lactation. Neither insufficient milk as a reason for not breastfeeding nor having received medication to stop milk flow was associated with increased risk. These findings are in accordance with accumulating evidence that lactation may have a weak protective effect on breast cancer risk. Am J Epidemiol 1997;146:932-8. breast feeding; breast neoplasms; lactation

Downloaded from http://aje.oxfordjournals.org/ by guest on December 12, 2011

While there are a number of reports that lactation, particularly prolonged lactation, may have a weak protective effect against breast cancer (1-17), the epidemiologic findings are far from conclusive. There are some indications that lactation may be protective only for premenopausal breast cancer (6, 11, 13, 15, 16), that early age at first lactation may be of significance (13, 18), that inability to breastfeed may be the relevant variable (6), and that duration of lactation is of importance (4-9, 11, 13-16). The potential role of lactation in the prevention of breast cancer is certainly of interest because it is a modifiable risk factor, and understanding of the role of lactation could contribute to our understanding of the etiology of a disease with enormous public health significance. In this casecontrol study of women in western New York, we examined the association of lifetime lactation with risk of breast cancer.
Received for publication January 17, 1997, and in final form July 15, 1997. 1 Department of Social and Preventive Medicine, State University of New York at Buffalo, Buffalo, NY. 2 Arizona Cancer Center, Tucson, AZ. 3 Department of Surgery, State University of New York at Buffalo, Buffalo, NY. Reprint requests to Dr. Jo L Freudenheim, Department of Social and Preventive Medicine, 270 Farber Hall, Buffalo, NY 14214. This paper is solely the responsibility of the authors and does not necessarily reflect the views of the National Cancer Institute.

MATERIALS AND METHODS

Lactation was examined in relation to breast cancer risk as part of a case-control study of breast cancer in premenopausal and postmenopausal women in western New York State. All participants provided written informed consent; procedures for protection of human subjects in this study were approved by the Human Subjects Review Board of the State University of New York School of Medicine and Biomedical Sciences and the review boards of each of the participating hospitals. Included were women between ages 40 and 85 years who were residents of Erie and Niagara counties, able to speak English, and in sufficiently good health to be interviewed. All were Caucasian. Women were considered to be premenopausal if they were currently menstruating or, if they were not menstruating because of a hysterectomy or other medical intervention, if they had at least one of their ovaries and were less than age 50 years. All other women were considered to be postmenopausal. Breast cancer cases were women with incident, pri-. mary, histologically confirmed breast cancer identified from pathology records of all of the major hospitals in the two counties during the period between November 1986 and April 1991. After identification of each case, her physician was contacted to obtain consent to invite her for an interview. In all, 66 percent of eligible premenopausal cases and 54 percent of eligible post932

Lactation and Breast Cancer Risk

933

menopausal cases agreed to participate. The primary reason for failure to participate was refusal by physicians to allow us to contact their patients. Among premenopausal women, 74 percent of nonparticipation was because of physician refusal; among postmenopausal women, it accounted for 71 percent The time between diagnosis and interview was 2 months on average. Controls were randomly selected from the two counties, frequency matched on age and county. Women under age 65 years were drawn from the New York State list of licensed drivers, and those age 65 and over were chosen from the list of the Health Care Finance Administration. Of eligible controls, 62 percent of the premenopausal and 44 percent of the postmenopausal women agreed to participate. Because the controls under age 65 years were all licensed drivers, we asked the cases in that age group if they had driver's Licenses; nine did not. These women were slightly less educated and slightly, although not significantly, older than the other cases. They were included in these analyses.
Interviews

pausal and postmenopausal breast cancer may differ, particularly with regard to lactation, analyses were stratified on menopausal status. Means and standard deviations of relevant variables were calculated; means for cases and controls were compared by using the Student t test (19). Odds ratios and 95 percent confidence intervals were calculated by using unconditional logistic regression (20). Tests for trend were computed from the p value of the logistic regression for the continuous variable with the appropriate covariates. For analyses of categorical data, risk was calculated relative to the indicated referent category. Adjusted analyses included control for age, education, family history of breast cancer, history of benign breast disease, age at menarche, age at first birth, parity, and, for postmenopausal women, age at menopause.
RESULTS

Downloaded from http://aje.oxfordjournals.org/ by guest on December 12, 2011

Participants were interviewed in their homes by trained interviewers. The entire interview lasted 2 hours on average. The interview focused on usual diet, medical history, family history of cancer, reproductive history, and other factors relevant to breast cancer risk. Height and weight were by self-report; body mass index was calculated as weight (kg)/height (m2). Family history of breast cancer was defined as having at least one first-degree relative (mother, sister, or daughter) with breast cancer. As part of the reproductive history, women were asked whether they had breastfed each of their infants and, if so, the length of time of lactation and the reason for stopping. For infants whom they did not breastfeed, the women were asked if they had received a shot or pill to stop milk flow. For analyses of risk associated with stopping lactation because of insufficient milk, we restricted this category to women who reported that they had stopped because of insufficient milk and who had lactated for less than one and a half months. There was no information collected regarding introduction of other foods to the infants or whether the infant was exclusively breastfed.
Statistical analysis

Characteristics of the cases and controls included in this study are shown in table 1. Cases and controls were of similar ages and education; parity was slightly higher for the controls than for the cases in the premenopausal women among this group of women with at least one livebirth. Secular trends with respect to lactation are evident in these data: About 45 percent of
TABLE 1. Characteristics of oases and controls, WMtorn Now York, 1987-1991
Caaesf Mean (SD) Controbt Mean (SO)

Premenopausal Age (years) Education (years) No. of livebirths Months of lactation No. with lactation experience (%) Ever Never 45.81 13.55 Z62 3.52 (3.99) (2.65) (1.29) (8.00) 46.21 13.73 Z86 5.20 (3.54) (2.46) (1.28)* (10.50)*

95 158

(37.5) (62.5)

119 147

(44.7) (55.3)

Postmenopausal Age (years) Education (years) No. of Bvebirths Months of lactation No. with lactation experience (%) Ever Never 62.90 12.28 3.22 4.71 (7.26) (2.70) (1.77)
(10.06)

63.22 12.09 3.27 5.22

(7.53) (2-51) (1.76) (9.22)

194 173

(52.9) (47.1)

245 182

(57.4) (4Z6)

To study the effect of lactation on risk independent of the well-documented effects of pregnancy, all analyses were limited to women with at least one livebirth. Because of indications that risk factors for premenoAm J Epidemiol Vol. 146, No. 11, 1997

* p < 0.05 for ttest comparison of means of cases and controls. f Data shown here are restricted to women with at least one livebirth. All values are means (standard deviations (SD)) except for the last variable, lactation experience. For that variable, values are the number of individuals (percent of the group).

934

Freudenheim et a).

the premenopausal and 57 percent of the postmenopausal controls ever breastfed. Mean duration of lactation was significantly greater in controls than in cases for the premenopausal women. Because we were concerned about selection bias, we examined the correlation between duration of lactation and other breast cancer risk factors among the controls. Correlations of these factors with duration of lactation were generally low, at least among the women whom we interviewed. Correlations were on the order of 0.3 or lower, with most lower than 0.1. The strongest correlate with duration of lactation was parity; the correlation was about 0.3 for both premenopausal and postmenopausal controls. The correlation of duration of lactation with education was 0.01 and 0.01 for the premenopausal and postmenopausal controls, respectively (data not shown). Risks associated with ever lactating and duration of lactation are shown in table 2. While there was a trend toward decreased risk associated with ever lactating for premenopausal women, 95 percent confidence intervals included the null. The odds ratio for 20 months
TABLE 2.

or more of lactation in the lifetime was 0.50 (95 percent confidence interval 0.21-1.12). There was some evidence of a trend toward a greater protective effect associated with increasing duration of lactation for the premenopausal women (p = 0.07). For postmenopausal women, there was no evidence of a protective effect of lactation on risk. We were concerned that it might be inappropriate to adjust for a history of benign breast disease in that such a history might be in the causal pathway. However, results were not different when this factor was removed from the adjusting variables. Results were also unchanged when dietary intake of vegetables and fruits was included in the model. Analysis was also done examining the effect of duration of breastfeeding the first child. For premenopausal women, again there was weak evidence of a protective effect, although it was not as strong as for total duration. The odds ratio for at least 6 months of breastfeeding the first child was 0.84 (95 percent CI 0.71-1.99), and the p for trend was 0.29. Among postmenopausal women, duration of lactation for the first child was not related to risk (data not shown).

Downloaded from http://aje.oxfordjournals.org/ by guest on December 12, 2011

Lactation and riek of breast cancer among parous women, western New York, 1987-1991
No. 04 cases No. of controls

Crude OR*

Adjusted ORt

96% CI*

Promenopausal Never Ever Duration (months)


0 <4 158 36 30 18 11 158 95 147 119 1.0 0.74 1.0 0.80

0.54-1.17

147
42 38

4-11 12-20
>2O

17
22

1.00 0.80 0.73 0.98 0.46

1.00 0.85 0.72 1.08 0.50

0.5O-1.42 0.41-1.27 0.78-1.96 0.21-1.12

Total

253

266 0.07 Postmonopausal

p for trend

Never Ever Duration (months)


0 <4

173 194

182 245

1.0

1.0

0.83

0.88

0.65-1.20

4-11 12-20
>20

173 89 57 22 26 367

182 104 74 37 30 427

1.00 0.90 0.81 0.62 0.91

1.00 0.96 0.88 0.64 1.08

0.66-1.40 0.57-1.38 0.56-1.34 0.79-1.87

Total p for trend

0.44

* OR, odds ratio; CI, confidence interval. f Adjusted for age, education, family history of breast cancer, history of benign breast disease, age at menarche, parity, age at first birth, body mass index, and age at menopause (postmenopausal women only).

Am J Epidemiol

Vol. 146, No. 11, 1997

Lactation and Breast Cancer Risk

935

As shown in table 3, there was no evidence of a difference in risk associated with age at first lactation for premenopausal women. For postmenopausal women, it appeared that the reduction in risk associated with lactation was limited to women who first lactated before age 25 years. However, while these results were adjusted for age at first birth, age at first birth and age at first lactation were highly correlated (r = 0.8), and the number of women with an early first birth who began lactation late was small (n = 24 and 21 for the premenopausal and postmenopausal women, respectively). We also examined the effect of age at first lactation within two categories of total duration of lactation (data not shown). For the postmenopausal women, the results were similar to those shown in tables 2 and 3. There was a some effect for age at first lactation but not for duration of lactation. There was no evidence that women who reported that they had stopped lactation because of a perception of insufficient milk were at subsequently greater risk of breast cancer (table 4). For these analyses, insufficient milk was defined as breastfeeding of an infant for less than 1.5 months. There was also no evidence that use of shots or pills to stop lactation increased breast cancer risk (table 5). In fact, in these data, the use of the medications appeared to be somewhat protective. When the analyses shown in table 5 were restricted to women who had never breastfed, there was again no evidence of increased risk with use of shots or pills (data not shown).

DISCUSSION

In this case-control study of women in western New York State, we found weak evidence of a protective effect of lactation on subsequent risk of premenopausal breast cancer. The reduction in risk for ever breastfeeding was about 20 percent for premenopausal and about 12 percent for postmenopausal women; for 20 or more months of lactation in the lifetime, there was a reduction in risk of about 50 percent for premenopausal breast cancer. There was some evidence of a linear trend with increasing duration of lactation for premenopausal but not for postmenopausal women. For premenopausal women, the effect was found primarily in the highest category of duration. While several epidemiologic studies give no evidence of an effect of lactation on risk of breast cancer (18, 21-30), numerous others, originating from a variety of geographic locations, report some reduction in risk, albeit generally a small one (1-17). Although in several cases, confidence intervals for results include the null, there is some consistency in the magnitude of the risk reduction found; there are a number of reports of a reduction in risk on the order of 20-50 percent for ever having lactated (8-17). In several studies, the effect of lactation was limited to premenopausal women or at least stronger among them than among postmenopausal women (6, 11, 13, 15, 16). It may be that the failure to see an effect in some of the studies with null results may be in part because premenopausal and postmenopausal women were examined

Downloaded from http://aje.oxfordjournals.org/ by guest on December 12, 2011

TABLE 3. Age at first lactation and risk of breast cancer among parous women, western New York, 1987-1991
Ago al first

lactation (years)

No. of cases

No. ol controls

Crude OR*

Adjusted ORt

95%CI

PmmenopausaJ Never lactated <25 225 Total p for trend Postmenopausal Never lactated <25 225 Total p for trend 175 104 88 367 183 168 76 427 0.52 1.00 0.65 1.00 0.67 1.21 158 50 45
253

147 69 50
266

1.00 0.67 0.84

1.00 0.80 0.72

0.51-1.26 0.42-1.23

0.13

1.21

0.46-0.95 0.80-1.83

* OR, odds ratio; Q, confidence interval. t Adjusted for age, education, family history of breast cancer, history of benign breast disease, age at menarche, parity, age at first birth, body mass index, total duration of lactation, and age at menopause (postmenopausal women only).

Am J Epidemiol

Vol. 146, No. 11, 1997

936

Freudenheim et al.
TABLE 4. Report of insufficient milk for any birth and risk of breast cancer among woman who ever lactated, western New York, 1987-1991
Insufficient milk No.tf cases No. of controls Crude OR' Adjusted ORt 95%CI

Pramonopausal Did not experience Experienced Total*


70 11 81

97 13 110

1.00 1.14

1.00 0.89

0.33-2.37

Postmenopausel Did not experience Experienced Total


110 30 140

129 51 180

1.00 0.71

1.00 0.63

0.34-1.16

* OR, odds ratio; Cl, confidence interval. t Adjusted for age, education, family history of breast cancer, history of benign breast disease, age at menarche, parity, age at first birth, body mass index, total duration of lactation, and age at menopause (postmenopausal women only). i Sample size is somewhat reduced because of missing data

Downloaded from http://aje.oxfordjournals.org/ by guest on December 12, 2011

TABLE 5. Uae of medication to stop lactation and risk of breast cancer among parous women, weetem New York, 1987-1991
No. of cases No. of controls Crude OR* Adjusted ORt 95% Cl<

Premenopausal Never Ever Total* 145 93 238 149 116 265 Postmenopausel Never Ever Total 226 106 332 271 135
406

1.00 0.82

1.00 0.71

0.41-0.96

1.00 0.94

1.00 0.89

0.62-1.29

OR, odds ratio; Cl, confidence interval. t Adjusted for age, education, family history of breast cancer, history of benign breast disease, age at menarche, parity, age at first birth, body mass index, total duration of lactation, and age at menopause (postmenopausal women only). $ Sample size is somewhat reduced because of missing data.

together. In our study, the magnitude of the reduction in risk was evident primarily for premenopausal breast cancer. It may also be that in some studies, particularly in industrialized populations, duration of lactation is insufficient to see an effect on risk. In our data, the protective effect was limited to women who breastfed for at least 20 months. When results from various populations are compared, differences in practices regarding infant feeding may need to be better understood to come to any conclusions regarding the effect of lactation on risk. In some of the populations studied, there were relatively few women who had never lactated (22, 28). When the reference group is parous women who have never

lactated, these women may have very different characteristics if they come from a population in which long periods of breastfeeding are the norm compared with coming from a population in which breastfeeding is less common. There are also differences between populations and secular trends within populations in infant feeding practices. These differences include the frequency of breastfeeding an infant and the manner of introduction of other foods and therefore the quantity of breast milk provided, such that the meaning of ever having lactated varies considerably. These differences, which may or may not be significant in terms of a biologic mechanism explaining any effect of lactation on breast cancer risk, have not been examined in detail. Brinton et al. (18) recently studied not only the total period of breastfeeding but also the time until supplementation of breast milk with formula. Their essentially null results were similar for both variables in a group of women with a low frequency of breastfeeding and breastfeeding of short duration. Nevertheless, there could be important differences in the effect on the breast tissue of frequent, exclusive breastfeeding of a infant and the effect of infrequent feeding of an older child who is primarily nourished by other foods. When months of lactation are computed, these are combined. In our study, we found some evidence that the protective effect of lactation for postmenopausal women was restricted to those with First lactation before age 25 years; effects did not differ much by age at first lactation for premenopausal women. However, because age at first birth was highly correlated with the age when lactation began, it was difficult in our data to distinguish the effects of these two factors. There have been two other reports that age at first
Am J Epidemiol Vol. 146, No. 11, 1997

Lactation and Breast Cancer Risk

937

lactation modified risk; however, in both of these, the modifying effect of age at first lactation was found for premenopausal women (13, 18). Others have also reported the strongest effect of duration to be a protective effect associated with duration of lactation of the first birth (8, 10). That effect may also relate to an effect of age at first lactation. We did not find a stronger effect associated with duration of lactation for the first birth, nor did we find that the effect of age at first lactation among postmenopausal women differed for those with longer or shorter durations of total lactation. One possible mechanism proposed by Byers et al. (6) for the apparently protective effect of lactation is that women who are unable to breastfeed their children because of insufficient milk supply are at greater risk and that the risk is related to malfunctioning of the breast tissue. They found risk to be greater among women who reported that they had ceased lactation because of insufficient milk. We found no such association nor did others (10, 13, 18). It has also been suggested that risk may be increased among women who do not breastfeed because of their exposure to medication to stop milk flow at the time of the birth of their infant. There was no evidence of an increase in risk in our study or in others (7, 13, 14, 18, 24) among women who reported that they had received such medications. There are several hypotheses for the mechanism of an effect of lactation on breast cancer risk. One possible mechanism is that pregnancy and lactation may induce differentiation of the breast tissue, making it more resistant to carcinogenesis (31). This possible mechanism implies that lactation at a younger age would protect against the effects of subsequent carcinogenic exposures and would therefore have a greater impact. As discussed above, there is indeed some evidence for such a protective effect. In addition, one would expect that women with an early age at first birth and at first lactation would show less risk with exposure to other carcinogens. Examination of such effect modification is merited but would require analysis of large groups of cases and controls. Another possible mechanism is that, because lactation is associated with amenorrhea, it is the associated reduced exposure to the cyclic hormones of menstruation that is significant. To study such an effect, detail would be required regarding women's experience of amenorrhea during lactation; the time of lactation-related anovulation is not necessarily the same as the time of lactation. However, one study that examined this factor found that lactation was more strongly associated with risk than was the duration of the amenorrhea (7). Further study of such a relation is warranted. Another potential
Am J Epidemiol Vol. 146, No. 11, 1997

mechanism for a protective effect of lactation is that the breast milk in some way protects the breast tissue. For example, there could be protection by excretion of fat-soluble carcinogens (32). There is some evidence that the level of a potential carcinogen, cholesterol /3-epoxide, is lower in the breast fluid of women during lactation and for up to 2 years after lactation (33). There are also lower levels of estrogen in breast fluid in the years after lactation (34). In one small study, there was evidence of higher levels of carotenoids in colostrum during lactation for subsequent births compared with first births (35). Lactation may alter the breast so that the tissue is less exposed to potential carcinogens and more exposed to potentially protective agents. In interpretation of the results of our study, the issues of bias relevant to case-control studies need to be taken into account. While the study was designed to be representative of the populations of the two counties, the response rate for this study was low. Most of the lack of case response was related to physician refusal to allow us to interview their patients; this refusal was probably not related to patient characteristics, although this possibility cannot be ruled out. It may be, however, that women with the most severe disease were not included so that our results may not generalize to that group. For controls, we conducted a study asking a few questions of potential participants over the telephone at the time of the initial contact. These questions were asked of both women who agreed to participate and those who did not. The questions were primarily with regard to usual diet, and data were not available on lactation or on descriptive factors such as education. There was strong agreement between participants and nonparticipants, providing some evidence that they may not differ (36, 37). Further duration of lactation was not correlated, at least among participants, with education among controls suggesting that if there were differential participation by socioeconomic status, it might not affect the results. Nevertheless, the possibility of selection bias among both cases and controls needs to be taken into account in interpretation of these findings. Recall bias and random error in recall are also concerns, in that lactation histories were by self-report. There is some evidence that reproductive histories are reported with good reliability by both cases and controls (38). While the evidence is certainly not conclusive, the consistent finding in a fair number of epidemiologic reports of a weak protective effect of breastfeeding certainly merits further consideration in terms of the potential mechanisms. Further examination of the breastfeeding experience, of longer durations of lactation, of the changes in breast tissue with breastfeeding,

Downloaded from http://aje.oxfordjournals.org/ by guest on December 12, 2011

938

Freudenheim et al. 16. Katsouyanni K, Lipworth L, Trichopoulou A, et al. A casecontrol study of lactation and cancer of the breast. Br J Cancer 1996;73:814-18. 17. Wu, AH, Ziegler RG, Pike MC, et al. Menstrual and reproductive factors and risk of breast cancer in Asian-Americans. Br J Cancer 1996;73:680-6. 18. Brinton LA, Potischman NA, Swanson CA, et al. Breastfeeding and breast cancer risk. Cancer Causes Control 1995;6: 199-208. 19. Snedecor GW, Cochran WG. Statistical methods. Ames, IA: The Iowa State University Press, 1980. 20. Breslow NE, Day NE, eds. Statistical methods in cancer research. Vol. I. The analysis of case-control studies. Lyon, France: International Agency for Research on Cancer, 1980. (IARC scientific publication no. 32). 21. Wynder EL, MacComack FA, Stellman SD. The epidemiology of breast cancer in 785 United States Caucasian women. Cancer 1978;41:2341-54. 22. Diekwaba FN. Breast cancer in Nigerian women. Br J Surg 1992;79:771-5. 23. Laing AE, Demenais FM, Williams R, et al. Breast cancer risk factors in African-American women: The Howard University Tumor Registry experience. J Natl Med Assoc 1993;85:931-9. 24. Thomas DB, Noonan EA. Breast cancer and prolonged lactation. The WHO Collaborative Study of Neoplasia and Steroid Contraceptives. Int J Epidemiol 1993;22:619-26. 25. London SJ, Colditz GA, Stampfer MJ, et al. Lactation and risk of breast cancer in a cohort of US women. Am J Epidemiol 1990; 132:17-26. 26. Brinton LA, Hoover R, Fraumeni JF Jr. Reproductive factors in the aetiology of breast cancer. Br J Cancer 1983;47: 757-62. 27. Brignone G, Cusimano R, Dardanoni G, et al. A case-control study on breast cancer risk factors in a southern European population. Int J Epidemiol 1987;16:35661. 28. Yuasa S, MacMahon B. Lactation and reproductive histories of breast cancer patients in Tokyo, Japan. Bull World Health Organ 1970,42:195-204. 29. MacMahon B, Purde M, Cramer D, et al. Association of breast cancer risk with age at first and subsequent births: a study in the population of the Estonian Republic. J Natl Cancer Inst 1982;69:1035-8. 30. Michels KB, Willett WC, Rosner BA, et al. Prospective assessment of breastfeeding and breast cancer incidence among 89,887 women. Lancet 1996;347:431-6. 31. Russo J, Russo IH. The etiopathogenesis of breast cancer prevention. Cancer Lett 1995;90:81-9. 32. Jensen AA. Levels and trends of environmental contaminants in human milk. In: Jensen AA, ed. Chemical contaminants in human milk. Boca Raton, FL: CRC Press, 1991:199-207. 33. Gruenke LD, Wrensch MR, Petrakis NL, et al. Breast fluid cholesterol and cholesterol epoxides: relationship to breast cancer risk factors and other characteristics. Cancer Res 1987; 47:5483-7. 34. Petrakis NL, Wrensch MR, Ernster VL, et al. Influence of pregnancy and lactation on serum and breast fluid estrogen levels: implications for breast cancer risk. Int J Cancer 1987; 40:587-91. 35. Patton S, Canfield LM, Huston GE, et al. Carotenoids of human colostrum. Lipids 1990;25:159-65. 36. Graham S, Hellmann R, Marshall J, et al. Nutritional epidemiology of postmenopausal breast cancer in western New York. Am J Epidemiol 1991; 134:552-66. 37. Freudenheim JL, Marshall JR, Vena JE, et al. Premenopausal breast cancer risk and intake of vegetables, fruits and related nutrients. J Natl Cancer Inst 1996;88:34O-8. 38. Paganini-Hill A, Ross RK. Reliability of recall of drug usage and other health-related information. Am J Epidemiol 1982; 116:114-22.

and of changes in the breast fluid with lactation will help to elucidate the role of this practice in the etiology of breast cancer. In any case, the advantages of breastfeeding to the health of an infant are well established and prompt a clear public health message that it ought to be encouraged.

ACKNOWLEDGMENTS

Supported in part by Public Health Service (PHS) grant CA11535 from the National Cancer Institute. Dr. Freudenheim is the recipient of PHS Research Career Development Award CA01633 from the National Cancer Institute.

Downloaded from http://aje.oxfordjournals.org/ by guest on December 12, 2011

REFERENCES 1. Yuan JM, Yu MC, Ross RK, et al. Risk factors for breast cancer in Chinese women in Shanghai. Cancer Res 1988;48: 1949-53. 2. Wang QS, Ross RK, Yu MC, et al. A case-control study of breast cancer in Tianjin, China. Cancer Epidemiol Biomarkers Prev 1992; 1:435-9. 3. Katsouyanni K, Trichopoulos D, Boyle P, et al. Diet and breast cancer: a case-control study in Greece. Int J Cancer 1986;38:815-20. 4. Rosero-Bixby L, Oberle MW, Lee NC. Reproductive history and breast cancer in a population of high fertility, Costa Rica, 1984-85. Int J Cancer 1987;40:747-54. 5. Layde PM, Webster LA, Baughman AL, et al. The independent associations of parity, age at first full term pregnancy, and duration of breastfeeding with the risk of breast cancer. Cancer and Steroid Hormone Study Group. J Clin Epidemiol 1989;42:963-73. 6. Byers T, Graham S, Rzepka T, et al. Lactation and breast cancer. Evidence for a negative association in premenopausal women. Am J Epidemiol 1985,121:664-74. 7. Chilvers C. Breastfeeding and risk of breast cancer in young women: United Kingdom National Case Control Study Group. BrMed J 1993;307:17-20. 8. Romieu I, Hernandez-Avila M, Lazcano E, et al. Breast cancer and lactation history in Mexican women. Am J Epidemiol 1996;143:543-52. 9. Tao SC, Yu MC, Ross RK, et al. Risk factors for breast cancer in Chinese women of Beijing. Int J Cancer 1988;42:495-8. 10. Siskind V, Schofield F, Rice D, et al. Breast cancer and breastfeeding: results from an Australian case-control study. Am J Epidemiol 1989;130:229-36. 11. Yoo KY, Tajima K, Kuroishi T, et al. Independent protective effect of lactation against breast cancer: a case-control study in Japan. Am J Epidemiol 1992;135:726-33. 12. Lubin JH, Burns PE, Blot WJ, et al. Risk factors for breast cancer in women in northern Alberta, Canada, as related to age at diagnosis. J Natl Cancer Inst 1982;68:211-17. 13. Newcomb PA, Storer BE, Longnecker MP. Lactation and a reduced risk of prernenopausal breast cancer. N Engl J Med 1994;330:81-7. 14. Yang CP, Weiss NS, Band PR, et al. History of lactation and breast cancer risk. Am J Epidemiol 1993; 138:1050-6. 15. McTiernan A, Thomas DB. Evidence for a protective effect of lactation on risk of breast cancer in young women. Results from a case-control study. Am J Epidemiol 1986; 124:353-8.

Am J Epidemiol

Vol. 146, No. 11, 1997

Das könnte Ihnen auch gefallen